“Bad-sensitive” Bayesian network trust model in Virtual Network

Network Virtualization is a promising concept to diversify the Future Internet architecture into separate Virtual Networks (VN) that can support simultaneously multiple services and architectures over a shared substrate network. This paper studies the trust evaluation of the VNs and the services provided by VNs. In particular, we propose a trust model based on the nai¿ve Bayesian network, which can represent the relationship between VN and the services. We also introduce the longevity factor to combine the old and current evaluation to make the model more objective. What´s more, we improve the model by the ¿bad-sensitive¿ process, which has several good characters, such as ¿rise-slowly, decline-quickly¿. Therefore, the improved model is more cautious and sensitive to the bad ratings. In our model, all parameters can be adjusted to make the model accommodate the different applications. Finally, the experimental results show that the improved model is self-adaptive and more desirable.